Some Statistics Mean Nothing

When I was in college, my major was statistics. It's not that I enjoyed it necessarily, it's more that it made sense to me. It was logical, understandable and all it required for me was a hefty amount of memorization; the math was the easy part. Your typical statistics class will teach you more about the different ways of computing numbers based on some analysis or study and how it can be broken down, as opposed to just talking about athletes' performances or simple poll results. Sure, those are derived from the statistical process, but they are more for news media to overemphasize to scare, warn or simply inform the general public. I'm here to tell you that some of these stats that they come out with are useless. They are pure garbage, meant only to make you believe there is a dire need to focus on them. You probably hear them more often than you realize, too.

A very serious poll, obviously.

As I've mentioned before, I work in an environment that houses many risks, including but not limited to cutting hazards, hazmat spills, lift equipment accidents and several other everyday tasks that could effect someone's long-term well-being. OSHA, the Occupational Safety and Health Administration, has certain standards that are federally upheld that are to be followed by all businesses with no exception. There are incidents at work that are rather minor, then there are some that are severe enough to be reported to OSHA. No matter what, each incident is recorded, including who was involved, what happened, when, where and why, plus how it could have been prevented. A lot of those facts documented are very important. It helps in guiding employees to working safer, such as knowing what hazards to avoid and how to prevent them.

Here's something that we hear about more frequent than anything else though: which day of the week most of our incidents have happened on. At most workplaces, it's Friday. End of the week, you're tired, you want to go home, and you end up losing focus. It makes sense. At my job, it just happens to be Tuesday. Does it make much sense? No. Does my manager try to rationalize why it's most frequent on Tuesdays? He does try. It's nothing more than mere circumstance. Accidents happen every day in several different lines of work, so it's no surprise to me that each workplace would have a different day observed as the one with the most incidents. When I go to work Tuesday, I'm actually going to hear about how it's more important to stay safe that day than the rest of the week...but we should still focus on safety on the other days as well. This is easily one of the most overstressed bullshit statistics I've ever seen or heard. I think it's time to put more emphasis on the physical environment and risks involved rather than what day of the week it just happened to fall on. Given I have weekends off, I'd say the day it fell on would be relevant if it was recorded on a Saturday. Otherwise, I don't want to hear about it. It's not important, period.

Good thing this happened on a Thursday.

He would have thrown those numbers way off.

How about when you're sitting and watching a game on t.v. and the commentators are going over the numbers of the players and teams that are playing against each other? Since baseball is quickly approaching the postseason, I'll draw my example from that. I previously wrote about Mike Trout, the rookie center fielder for the Angels, who is having a phenomenal season, and is still very close to being on pace with his projected numbers for the full season. When these broadcasters talk about his home run, hit, stolen base and RBI totals, they are standard mentions of a player's accomplishments in a week, month, season or even career. It's when they go into these elaborate, situational statistics that are of no use to anybody. Nobody cares who has the most home runs off of lefty pitchers in September in the seventh inning or later when their team is losing. Only the commentators are concerned about the player's career batting average with men in scoring position, two outs and a full count in night games against the Orioles. If someone finds that random/specific of a stat to be relevant, then they really aren't focused in on the situation at hand. Whatever overly narrowed down number they can produce has no bearing on what the batter is going to do right then and there. They need to stop making up these ridiculous qualifiers of what makes a stat significant or not.

After they deliver that meaningless, nonsense stat,

they will immediately lean in for a kiss. Sad but true.

The only other kind of stats that bother me are the polls that come out during some news story about what some large group of people think about any particular issue. For instance, the next time you see a poll come out about what 'Americans say' on some political issue, really take a minute to understand how generalized their claim is. In all of my years, aside from surveys I have volunteered to participate in, I have not once been approached by anyone asking me the same question I have seen them asking in these polls. Yet they claim that 'Americans say' whatever it is they say. Well sure, you're polling some Americans. If everyone in America was asked the same question, and then the results were released, that would be a very accurate display of what 'Americans say'. When you're only polling a sample, it's extremely simple for the individuals in control of the study to introduce bias to the mix. That throws off what would be the actual results if you got everyone's answer. Not everyone supports the right to gay marriage. I guarantee you that if someone polled every person as they are leaving church some Sunday morning, you're going to get some pretty convincingly one-sided opinions that don't necessarily reflect those of non-churchgoers. See how easy it is to provide the general public with a bullshit stat? Don't believe everything you hear.

They may get some interesting poll results from people leaving this building as well.

Mark Twain attributed a quote from Benjamin Disraeli, saying, "There are three kinds of lies: lies, damned lies and statistics." Clearly, the manipulation of numbers and figures was not uncommon back in Twain's time, nor was it in the 19th century when Disraeli was British Prime Minister. I'm just echoing their warning of sorts. A statistic is easy to twist to benefit one side of an issue over another. Next time you read up on one, whether it's an important one or not, just see how large-scale of a sampling they actually used to get that information. Don't take it for it's word as absolute truth. Sometimes, it couldn't be further from that.

D.